# Boundary layer flow and dispersion over isolated hills and valleys

## Overview of Tests

Full details of the experimental procedures are provided in Kurshudyan et al (1981). Here we summarise the summary given in CA. The upstream (simulated, neutrally stable atmospheric) boundary layer was developed using a combination of an inlet fence and distributed (10mm, sanspray) roughness. At the hill location (but in its absence), 8.7m downstream of the fence, the boundary layer was only slowly developing and had a depth of about 1m. Profiles of the mean velocity and turbulence stresses are included in the data set, along with all the data obtained for the three hill cases. This data set can be found in the ERCOFTAC database (case 69) and is the final smoothed data generated by Trombetti et al (1991), see Measured data. The free-stream velocity, U0, was 4m/s so that the Reynolds number based on hill height was about 3x104. The scalar source was a 15mm diameter porous sphere, which provided a neutrally buoyant, isokinetic release, with a continuously monitored (constant) flow rate, Q. Averaging times for velocities, stresses and concentrations were at least two minutes, with sampling rates high enough to ensure that statistical errors arising from a finite sample size were insignificant.

Velocity and turbulence profiles were obtained at numerous axial locations. For example, for the lowest slope hill (which had a total length at z=0, a=2L, of 16H or about 1872mm) profiles were taken at x/H=±16, ±8, ±4 and 0, i.e. at one hill length upstream and downstream from the summit, at the upwind and downwind base and half-way up and down the hill slopes. (Note that the x-coordinate origin is the z=0 position beneath the hill summit). Similar relative positions were used for the other hills. The source stack positions were xs/L=±1 or 0, and zs=0.25H, 0.5H, H and 1.5H for each axial location. For each stack position (xs, zs) and each hill (and also the no-hill case), surface concentration profiles were obtained to downwind fetches where the concentration had fallen by at least two orders of magnitude from its maximum. In addition, vertical profiles were obtained at the same x-locations as those used for the flow profiles and corresponding cross-stream profiles were also obtained (generally only at z=zs). The raw concentration values were normalised as: Cn=CU0H2/Q and it is these normalised data which are contained in the aforementioned database. There are no ‘global’ parameters (e.g. hill amplification factors) contained in the database. Most of the computational comparisons contained in CA consist of profile comparisons, although some of these are normalised – e.g. surface concentration normalised by its maximum value or by the corresponding maximum ground level concentration in the absence of the hill.

Table A Summary description of all test cases.
NAME GNDPs PDPs MPs
Re Hill aspect ratio Scalar location Detailed data DOAPs
EXP1 Hill flow and dispersion 3x104 3, 5, 8 Xs/L = ±1 or 0, and Zs = 0.25H, 0.5H, H and 1.5H C, U, stresses U/U0 C/C0 uw/U02, etc

Table B Summary description of all measured parameters and available data files. Tabulated data available are provided in the files listed in Tables C and D. Note that the flat surface reference case is in the file: Flatrefcase.dat
MP1 Normalised C MP2 ${\displaystyle U{(ms^{1})}}$ MP3 stresses DOAPs or other miscellaneous data
EXP 1 File names given in Table C As file names listed in Table D As file names in the form: hnms.dat where n is hill case (3, 5 or 8) and m is a two- digit file number Listed in Table C Specified in files listed in Table C Specified in files listed in Table C

 Data filename Data filename Data filename NO HILL Flatrefcase.dat HILL 3 HILL 5 HILL 8 h301s.dat h501s.dat h801s.dat h302s.dat h502s.dat h802s.dat h303s.dat h503s.dat h803s.dat h304s.dat h504s.dat h804s.dat h305s.dat h505s.dat h805s.dat h306s.dat h506s.dat h806s.dat h307s.dat h507s.dat h807s.dat h308s.dat h508s.dat h808s.dat h309s.dat h509s.dat h310s.dat h510s.dat h810s.dat h311s.dat h511s.dat h811s.dat h312s.dat h512s.dat h812s.dat h313s.dat h513s.dat h813s.dat h314s.dat h514s.dat h814s.dat h315s.dat h515s.dat h815s.dat h316s.dat h516s.dat h816s.dat h817s.dat

TABLE D List of available data files for normalised concentrations.
NO HILL HILL 3 HILL 5 HILL 8
Surface concentration for source heights 29, 59 etc., mm -
Flatref029G.dat Surface concentration for source heights at 29 or 59 mm, located at Upwind base (Uw), summit (Su), or Downwind base (Dw), of hill. Surface concentrations for source heights at 29mm and located at Upwind base (Uw), Summit (Su), or Downwind base (Dw), of hill. Surface concentrations for source heights at 29mm or 59mm and located at Upwind base (Uw), Summit (Su), or Downwind base (Dw) of hill.
Flatref059G.dat
Flatref117G.dat
Flatref176G.dat
Flatref234G.dat
Flatref351G.dat h8CUwS29G.dat
h3CUwS29G.dat h5CUwS29G.dat h8CSuS29G.dat
h3CSuS29G.dat h5CSuS29G.dat h8CDwS29G.dat
Vertical profiles, for 29 and 117mm source heights and given x mm. h3CDwS29G.dat h5CDwS29G.dat h8CDwS59G.dat
h3CDwS59G.dat
FlatrefC029S585V.dat
FlatrefC117S330V.dat
FlatrefC117S585V.dat
FlatrefC117S1170V.dat
FlatrefC117S1600V.dat
FlatrefC117S1872V.dat

## Test Cases - Further Details

Description of Experiment

This is given above.

Boundary Data

The upstream boundary layer data is included in the database. The surface roughness length, zo and normalised friction velocity, u*/Uo, in the usual notation, were 0.16mm and 0.047, respectively. Total turbulence kinetic energy was not measured (although individual profiles of axial and vertical stress and shear stress are available), so in the computations it was estimated – see below.

Measurement Errors

Errors in the hot wire data were reckoned to be within the usual limits obtainable with careful calibration (including proper yaw calibration) – i.e. ±2% on mean velocity and 10-15% in the stresses (normalised by the Uo2). The experimental programme included considerable effort to ensure as high an accuracy as possible. Concentration data was obtained using standard Flame Ionisation Detector techniques and the errors were estimated to be between 4 and 20%; consequent errors in the normalised data are similar, since the accuracy of monitoring both Q and Uo was high.

Measured Data

The original data is contained in text files as Case 69 on the ERCOFTAC database, at the web address: http://cfd.mace.manchester.ac.uk/ercoftac/. Access to these files is limited and will not be of much help, but they were used to construct the files listed in Tables C & D above. These include data on mean and turbulent velocities and scalar concentrations for three hills and with the various source locations specified earlier.

References

Busuoli, M., Trombetti, F. & Tampieri, F. (1993). Data sets for studies of flow and dispersion in complex terrain: I) the RUSVAL wind tunnel experiment (flow data). CNR Technical Report No.3, FISBAT-RT-93/1

Castro, I.P. & Apsley, D.D. (1997). Flow and dispersion over topography: a comparison between numerical and laboratory data for two-dimensional flows. Atmos. Env., 31, 839-850.

Khurshudyan, L.H and Snyder, W.H & Nekrasov, I.V (1981). Flow and dispersion of pollutants over two-dimensional hills. U.S. Env. Prot. Agcy. Rpt. No. EPA-600/4-81-067. Res. Tri. Pk., NC.

Khurshudyan, L.H and Snyder, W.H & Nekrasov, I.V, I.V and Lawson R.E., Thompson R.S. & Schiermeier F.A. (1990). Flow and dispersion of pollutants within two-dimensional valleys. U.S. Env. Prot. Agcy. Rpt. No. EPA-600/4-79-051. Res. Tri. Pk., NC.

Maurizi, A. (2000) Numerical simulation of turbulent flows over 2D valleys using three versions of the k-ε closure model. J. Wind Eng. Ind. Aero., 85, 59-73.

Trombetti, F., Martano, P. & Tampieri, F. (1991). Data sets for studies of flow and dispersion in complex terrain: I) the RUSHIL wind tunnel experiment (flow data). CNR Technical Report No.1, FISBAT-RT-91/1